• ETRI Journal
• /
• 제41권4호
• /
• pp.494-505
• /
• 2019

In this paper, we propose a new framework for detecting the unauthorized dumping of garbage in real-world surveillance camera. Although several action/behavior recognition methods have been investigated, these studies are hardly applicable to real-world scenarios because they are mainly focused on well-refined datasets. Because the dumping actions in the real-world take a variety of forms, building a new method to disclose the actions instead of exploiting previous approaches is a better strategy. We detected the dumping action by the change in relation between a person and the object being held by them. To find the person-held object of indefinite form, we used a background subtraction algorithm and human joint estimation. The person-held object was then tracked and the relation model between the joints and objects was built. Finally, the dumping action was detected through the voting-based decision module. In the experiments, we show the effectiveness of the proposed method by testing on real-world videos containing various dumping actions. In addition, the proposed framework is implemented in a real-time monitoring system through a fast online algorithm.

• 전자통신동향분석
• /
• 제35권3호
• /
• pp.20-33
• /
• 2020

Temporal action detection (TAD) in untrimmed videos is an important but a challenging problem in the field of computer vision and has gathered increasing interest recently. Although most studies on action in videos have addressed action recognition in trimmed videos, TAD methods are required to understand real-world untrimmed videos, including mostly background and some meaningful action instances belonging to multiple action classes. TAD is mainly composed of temporal action localization that generates temporal action proposals, such as single action and action recognition, which classifies action proposals into action classes. However, the task of generating temporal action proposals with accurate temporal boundaries is challenging in TAD. In this paper, we discuss TAD technologies that are considered high performance in terms of representative TAD studies based on deep learning. Further, we investigate evaluation methodologies for TAD, such as benchmark datasets and performance measures, and subsequently compare the performance of the discussed TAD models.

• 한방안이비인후피부과학회지
• /
• 제7권1호
• /
• pp.1-13
• /
• 1994

The Effects of Hoichunyanggyuksan on the Anti-allergic Effect, Analgesic Action, Anti-inflammatory Action and Antipyretic Action. Experimental studies were done to research the clinical effects of Hoichunyanggyuksan on the Anti-allergic effect, Analgesic action, Anti-inflammatory action and Antipyretic action. The results obtained as follows; 1. On vascular permeability responses to intradermal histamine, Hoichnyanggyuksan showed significant effect. 2. In the homologous PCA provoked by the IgE-like antibody against white egg albumin, Hoichunyanggyuksan showed the decreasing tendency, but was none significant effect. 3. In the delayed type hypersensitivity responses to Picryl chloride, Hoichunyanggyuksan was proved significant effect. 4. In the delayed type hypersensitivity resposes to SRBC, Hoichunyanggyuksan revealed significant effect. 5. In Anti-pyretic action by yeast method, Hoichunyanggyuksan showed significant effect. 6. In Anti-inflammatory action by carrageenin method, Hoichunyanggyuksan showed significant effect. 7. In analgesic action by acetic acid method, Hoichunyanggyuksan was recognized significantly.

• Asian Pacific Journal of Cancer Prevention
• /
• 제15권17호
• /
• pp.7343-7350
• /
• 2014

Cancer is one of the leading causes of death in Thailand as well as other countries. Evidences from studies in Asia show that alcohol consumption has been linked to an increased risk for various types of cancer. In the qualitative participatory action research presented here, the possibilities of a volunteer women's movement to reduce harmful alcohol consumption by facilitating the creation of emancipatory knowledge in the Thai-Isaan community was explored. During in-depth interviews with 10 female volunteer participants, the following themes emerged: merit making, developing a sense of personal empowerment, and the experience of participating in an action plan that utilized various strategies. The project activities empowered participants to create and share knowledge which was then applied toward action for change. Adoption of this type of program by local health care providers to reduce health problems and risks related to alcohol misuse in similar contexts and cultures is recommended.

• 생약학회지
• /
• 제10권2호
• /
• pp.73-77
• /
• 1979

Antiinflammatory action of phytolaccosides B and E isolated from the roots of Phytolacca americana was investigated and compared with that of aescin. Phytolaccoside B, when it was administered intravenously, exerted a potent inhibitory effect on the transudation caused by absorptive sponge pellet implanted s.c. in rats and on carrageenin induced edema in rats. Phytolaccoside E showed a weaker action than Phytolaccoside B did. Hemolytic potency of the saponins was decreased in the order of aescin, phytolaccosides B and E showing a same tendency as in antiinflammatory action. The experimental result suggests that the mode of action of phytolaccosides is the same as that of aescin.

• 약학회지
• /
• 제34권6호
• /
• pp.439-446
• /
• 1990

The general pharmacological actions of recombinant human growth hormone (rHGH) were investigated. It had hypothermic action but neither sedative nor analgesic action. No pharmacological effects were observed in isolated guinea pig ileum and tracheal muscle and rat fundus and uterus. Slight hypotensive action with no effect on respiration was revealed at a dose of 20 IU/kg i.v. of rHGH in rabbits. The rHGH exhibited a weak inhibitory action of glucose tolerance in normal rats, significantly lowered the blood glucose contents in adrenalectomized rats 20 min after i.v. administration (80IU/kg), and produced a significant inhibitory effect on in vitro glycerol release in epinephrine-stimulated epididymal fat pad segments of rats.

• 한국경영정보학회:학술대회논문집
• /
• 한국경영정보학회 2008년도 춘계학술대회
• /
• pp.428-433
• /
• 2008

The online advertising industry's business model undertakes the change from CPM (cost-per-mille)-based to CPC (cost-per-click)-based. However, due to the problem of 'Click Fraud', CPA (cost-per-action) has been regarded as a new step. For CPA, publishers need to get information after a user clicks an advertisement. Therefore, in CPA, the key is to get Conversion Action Data (CAD). This paper introduces two existing mechanisms for getting CAD, compare their characteristics, and analyze their limitations. Then the two new mechanisms are introduced and their requirements and feasibility are analyzed. Lastly, we compare the existing two and the new two mechanisms, and point out each mechanism's business possibility, value and Application Area. This paper will help publishers choose the most appropriate mechanism on the basis of their situation.

• Nuclear Engineering and Technology
• /
• 제29권5호
• /
• pp.368-374
• /
• 1997

This paper presents a new approach to the evaluation of an accident management strategy when an operator action is involved. This approach classifies the failure in implementing a given strategy into 4 possible mechanisms, and provides their corresponding quantification methods : 1) the failure to formulate correct intention by operators, 2) the failure to take an adequate action following a correct diagnosis, 3) the failure of a system operation following an adequate action, and 4) the failure due to a delayed action. The proposed method was applied to assess a cavity flooding strategy that uses containment spray system (CSS), and the result shows that the method is more appropriate in evaluating accident management strategies when human action is involved.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권5호
• /
• pp.2327-2347
• /
• 2018

Human actions can be recognized from depth sequences. In the proposed algorithm, we initially construct depth, motion maps (DMM) by projecting each depth frame onto three orthogonal Cartesian planes and add the motion energy for each view. The body part of the action (BPoA) is calculated by using bounding box with an optimal window size based on maximum spatial and temporal changes for each DMM. Furthermore, feature vector is constructed by using BPoA for each human action view. In this paper, we employed an ensemble based learning approach called Rotation Forest to recognize different actions Experimental results show that proposed method has significantly outperforms the state-of-the-art methods on Microsoft Research (MSR) Action 3D and MSR DailyActivity3D dataset.

• ETRI Journal
• /
• 제44권2호
• /
• pp.327-338
• /
• 2022

Classification models for human action recognition require robust features and large training sets for good generalization. However, data augmentation methods are employed for imbalanced training sets to achieve higher accuracy. These samples generated using data augmentation only reflect existing samples within the training set, their feature representations are less diverse and hence, contribute to less precise classification. This paper presents new data augmentation and action representation approaches to grow training sets. The proposed approach is based on two fundamental concepts: virtual video generation for augmentation and representation of the action videos through robust features. Virtual videos are generated from the motion history templates of action videos, which are convolved using a convolutional neural network, to generate deep features. Furthermore, by observing an objective function of the genetic algorithm, the spatiotemporal features of different samples are combined, to generate the representations of the virtual videos and then classified through an extreme learning machine classifier on MuHAVi-Uncut, iXMAS, and IAVID-1 datasets.